Chapter 6

The Role of Classroom Discussion

Diego Carrasco and David Torres Irribarra

Abstract Past research has shown that students in schools with greater levels of
open classroom discussion, have more positive attitudes toward other groups and
hold more democratic attitudes. Students do not learn citizenry only by knowledge
acquisition; school practices such as classroom discussion foster critical thinking,
help students to understand others and reduce closed-mindedness. Students with a
higher exposure to classroom discussion were hypothesized to display more tolerant
attitudes to other groups and hold more egalitarian values in general. The analytical
strategy in this chapter uses a three-level path analysis with support for equal rights
for women, for all ethnic/racial groups and for immigrants as outcomes.
Appropriate variable centering and random intercepts for schools and countries
enabled relationships between classroom discussion and the outcomes to be
determined. Open classroom discussion was found to be positively related to
egalitarian values across all samples, accounting for 5 to 8% of school variance,
depending on the outcome.

Keywords Attitudes toward diversity International Civic and Citizenship


Education Study (ICCS) International large-scale assessments


Multilevel path analysis Open classroom for discussion

6.1 Introduction

One of the main aims of civic education is the promotion of democratic values,
through the promotion of civic knowledge and the endorsement of democratic
attitudes (Lenzi et al. 2014). The interpretation of democracy as “a mode of

D. Carrasco (&)
Centro de Medición MIDE UC, Pontificia Universidad Católica de Chile,
Santiago, Chile
e-mail: dacarras@uc.cl
D. Torres Irribarra
Escuela de Psicología, Pontificia Universidad Católica de Chile, Santiago, Chile

© International Association for the Evaluation 87

of Educational Achievement (IEA) 2018
A. Sandoval-Hernández et al. (eds.), Teaching Tolerance in a Globalized World,
IEA Research for Education 4, https://doi.org/10.1007/978-3-319-78692-6_6
88 D. Carrasco and D. Torres Irribarra

associated living” (Dewey 1916, p. 101) requires citizens to behave socially in

different contexts. Schools are a key scenario for the socialization of these different
modes of associated living.
The presence of injustice in its various forms erodes the legitimacy of
democratic institutions. Prejudice, corruption and a lack of commitment to equality
are primary concerns in this regard. Racism, sexism and anti-immigrant attitudes
are all examples of different forms of prejudice. In contrast, egalitarian attitudes are
the positive formulation of these dispositions. Because attitudes are developed and
learned, it is generally thought that these can be unlearned as well (Zick et al. 2011).
Schools are a major actor in this regard, as schools promote norms and values about
how students should act in their community and their nation (Quaynor 2012). Thus,
schools are an active agent in the process of supporting students to unlearn negative
intergroup attitudes and to promote egalitarian attitudes and other relevant
democratic values.
What schools do to promote democratic values matters? Past research has
highlighted the relevance of school environments within civic education research,
especially the perceptions of open classroom discussion, for its impact on different
citizenship outcomes. This includes its positive relation to civic knowledge (Schulz
2002; Schulz et al. 2010; Torney et al. 1975), its positive relation to tolerant
attitudes (Caro and Schulz 2012), and its negative relation to youth alienation
(Torney-Purta 2009), by which we mean adolescents with high political disaffection
and generalized negative attitudes toward others.
Measures of open classroom discussion aim to capture an aspect of the learning
environment expected to influence the development of democratic principles. The
open classroom discussion scores indicate whether students can discuss, during
regular lessons, political and social issues in their classrooms, what level of
encouragement they receive in developing informed opinions during those dis-
cussions, and if students receive teacher guidance to debate the arguments. Thus,
this score measures how regularly students can openly discuss political and social
issues at their school.
As open classroom discussion is a reflective measure of the learning
environment, and not an individual difference like socioeconomic background
(Lüdtke et al. 2008); care must be taken when using these responses as school
differences in multilevel models to avoid underestimating some of the effects
(Lüdtke et al. 2009). The present chapter relies on this approach, where student
responses are the source of information about their school practices and students
rate their learning environments.
After reviewing the research literature on civic education and attitudes toward
others, we developed a plausible link between the learning environment differences
and students’ endorsement of egalitarian attitudes. This reflective measure approach
to school climate factors informed our estimated model.
6 The Role of Classroom Discussion 89

6.2 Conceptual Background

6.2.1 Schools and Egalitarian Attitudes

When researchers study intergroup attitudes, they commonly find a relationship

between educational attainment and prejudice (Easterbrook et al. 2015). For
example, people with lower levels of education are generally more prone to prej-
udice than people with higher educational attainment (Coenders and Scheepers
2003). Moreover, longitudinal studies comparing academic tracks and vocational
tracks have found that students in academic tracks develop more tolerant attitudes
over time, while students on vocational tracks develop less tolerant attitudes
toward others (Hooghe et al. 2013a, b; Vollebergh 1996). Thus, different school
experiences may shape youth attitudes toward other groups.
How can these differences be explained? The ‘sophistication hypothesis’
(Highton 2009; Luskin 1990) suggests that people develop the necessary cognitive
skills for democracy through education. The schooling process provides more
sophisticated knowledge to people, and this information promotes the development
of less prejudiced attitudes (Easterbrook et al. 2015). Thus, schools which provide a
more democratic environment are expected to foster more egalitarian attitudes.
Complementary to this, within this framework, socially and economically
disadvantaged groups are thought to be more prone to prejudice (Lipset 1959)
because they are exposed to more negative experiences which often translate into
ethnic prejudice. Restrictions in cultural, intellectual or family resources prevent
low-status members of society from expanding their understanding of different
groups and ideas (Carvacho et al. 2013). In essence, differences in “cultural capital”
(the ability to understand the way of life of others; Houtman 2003) hinders the
development of egalitarian attitudes. Thus, students in schools that foster reflection
and the understanding of other perspectives are expected to display more positive
attitudes toward other social groups.
Creating opportunities for classroom discussion is an important way of fostering
understanding of alternative points of view, as a way of increasing cultural capital.
This is consistent with Dewey’s theories on education and democracy. Van der
Ploeg (2016, p. 148) put it thus:
For Dewey, morality is dependent on deliberation, reflection and insight. This means that
morality relies on communication and cooperation. For an adequate assessment of the
moral value of my actions, I need others’ contributions. Given that common good has to do
with the conditions underlying the self-development of everyone, and so those of others as
well, I require insight into others’ beliefs and wishes in order to contribute. The only way to
acquire this is by interaction and communication. In addition, my inquiry and reflection can
benefit from cooperation with others, for instance inquiring together, reflecting together,
benefiting from one another’s expertise, sharing knowledge, insight and experience and
having discussions. [Emphasis added]

In this sense, open classroom discussion can be understood theoretically as

creating a privileged opportunity to gain “insight into others’ beliefs and wishes”, as
90 D. Carrasco and D. Torres Irribarra

a school practice that fosters the understanding of others. A more psychological

account posits that educational interventions directed to reduce the “need for clo-
sure”, a form of cognitive conservatism, and closed-mindedness, might reduce
prejudice in an indirect way (Van Hiel et al. 2004). Differences between schools in
this respect may explain the endorsement of different egalitarian values between
schools.

6.2.2 Past Research

The importance of open classroom discussion in the development of social and

political attitudes has been extensively researched through the data collected by the
1999 Civic Education Study (CIVED) and International Civic and Citizenship
Education Study (ICCS) 2009 (see for example Barber et al. 2015; Campbell 2008;
Caro and Schulz 2012; Godfrey and Grayman 2014; Schulz 2002; Schulz et al.
2010; Torney-Purta 2009). While there is no consensus regarding the psychological
or social mechanisms through which open classroom discussion operates, these
studies have consistently backed its role as an explanatory factor in the develop-
ment of civic knowledge, a positive outlook toward political debate, and an interest
in informed voting (Campbell 2008; Godfrey and Grayman 2014).
Despite its frame of reference being the classroom, the responses of students in
open classroom discussion have been studied as differences in students’ experiences
(see Caro and Schulz 2012; Torney-Purta 2009), and as differences between
schools. In the latter approach, open classroom discussion has been assessed by
excluding students’ individual scores and using school means only (for example,
see Godfrey and Grayman 2014), or by including students’ individual scores and
school means at the same time (see Schulz 2002; Schulz et al. 2010), as in common
compositional models (Caro and Lenkeit 2012; Willms 2010).
As open classroom discussion scores are not a traditional individual difference
measure in the way that, for example, socioeconomic background is (Lüdtke et al.
2008), the traditional model specification for compositional effects may result in
unnecessary overcorrections of the between school difference (Lüdtke et al. 2009).
Thus, standard recommendations for centering individual scores and school means
scores to the overall mean (O’Connell and McCoach 2008) do not apply for these
measures in the same way and have negative consequences for the intended
inference.
Lüdtke et al. (2009) argued that the study of school environments should center
its attention on the between-school differences when students are the informants.
This translates into appropriately identifying if a measure is a reflective construct of
a cluster level (Stapleton et al. 2016), and using appropriate centering techniques
for responses. In practice, this treats student answers as if they are raters of their
own learning environment.
The present work aims to uncover the role of open classroom discussion by
measuring the between-school differences of open classroom discussion and using
6 The Role of Classroom Discussion 91

group mean centering where appropriate. Additionally, previous results in the

literature of open classroom discussion have reported a buffer effect over students’
disadvantaged background and other citizenship outcomes (for example Campbell
2008; Godfrey and Grayman 2014). In this chapter, we explore the plausible
moderating effect of open classroom discussion on student characteristics and
support for equal rights for women, all ethnic groups and immigrants.

6.3 Methods

6.3.1 Data

The data were taken from ICCS 2009 (for the specific description of this dataset see
Chap. 2 in this volume). The final sample used for the analyses included in this
chapter shows small variations from the original dataset, as the set of variables
involved in these analyses have specific missing patterns. The final sample was
140,650 students, 5369 schools and 38 countries.

6.3.2 Variables

Dependent Variables
The dependent variables were attitudes toward equal rights for disadvantaged
groups, including: immigrants, ethnic groups and women. These were derived from
the original items from the attitudes toward gender equality, equal rights for all
ethnic/racial groups and equal rights for immigrants that appeared originally in
ICCS 2009. Using a multi-group confirmatory analysis, factor scores were derived
and used as manifest variables. Thanks to reaching measurement invariance, these
outcomes were in a comparable scale (see Chap. 3 for more details). These three
variables were included in the analysis in this chapter, thus allowing us to account
for the distribution of these three factors together.
Independent Variables
As explanatory variables (Table 6.1), we used the following factors from the ICCS
2009 public data file: civic knowledge (PV1CIV-PV5CIV) plausible value scores
from students, open classroom discussion (OPDISC), socioeconomic status of the
students (NISB), gender (SGENDER), and immigrant status (IMMIG). The last was
recoded as a dummy variable, where the category of reference consisted of all
non-immigrant students, and the effect category consisted of all students with an
immigrant background, including students from a first generation immigrant
background and students born in a different country.
92 D. Carrasco and D. Torres Irribarra

Table 6.1 Independent variables from ICCS 2009

Variable name Independent Type Description
variables
PV1CIV-PV5CIV Civic Continuous Five plausible values stand for student
knowledge civic knowledge scores. These were
divided by the expected international
standard deviation (10 pts) of the scale
OPDISC Open classroom Continuous Open classroom discussion was
discussion decomposed into student deviations from
their school mean, and school means
within each country
NISB Socioeconomic Continuous Socioeconomic status was decomposed
status of the into student deviations from their school
students mean, and school means within each
country
SGENDER Student gender Dummy Female = 1, male = 0
IMMIG Immigrant Dummy Students with immigrant
status background = 1, native = 0

6.3.3 Analytical Strategy

We specified a three-level path analysis model, where support for equal rights for
women, support for equal rights for all ethnic/racial groups, and support for equal
rights for immigrants are included as response variables. This allowed us to inspect
the relationship between four variables of interest and an outcome while controlling
for the level of the other dependent variables. This model included random inter-
cepts at both school and country level, separating all observation dependencies and
allowing us to draw cluster-specific inferences for school learning environments
(McNeish et al. 2017). With the appropriate centering, this model supports the
estimation of the overall mean of our covariate of interest across all samples
(Brincks et al. 2017).
Open classroom discussion is a reflective measure of the school environment
(Lüdtke et al. 2008; Stapleton et al. 2016) and not a classical individual difference
measure. Its frame of reference is the learning environment and not just the
experience of students as individuals. As such, it allows the capturing of the
experience of students as a collective, relative to the learning environments students
are in. Thus, in order to appropriately study its relationship to our outcomes, we
divided this factor into two components: the within-cluster variation and the
between-cluster variation (as suggested by Campbell 2008). This was achieved by
centering the open classroom discussion scores to the school means. Additionally,
we wanted to collect the pooled regressions estimate of open classroom discussion
for the 38 samples included in this study. This provides an overall mean estimate of
this covariate, across all samples. Hence, to achieve this, we had to adjust the
previous between-cluster variation so it was correctly centered within countries (see
6 The Role of Classroom Discussion 93

Brincks 2012; Brincks et al. 2017). Using this specification, we can explain the
relationship between open classroom for discussion across all compared learning
environments and our three outcomes of interest in a single model.
We included as control variables: socioeconomic background of students, civic
knowledge scores, gender and immigrant background. The two first variables were
included in the model using the same centering approach as used for open class-
room discussion. This, enabled us to assess whether the main effect under study was
resistant to school differences across all samples in terms of the socioeconomic
composition of the schools and to civic knowledge levels of the schools. In contrast,
the last two variables were included purely as controls and were entered into the
model centered to the country overall means so as to remove their effects (Heck and
Thomas 2015). Hence, the estimates of the model accounted for school environ-
ments, with a similar composition in terms of gender and immigrant background.
To assess the impact of the open classroom discussion levels of schools, we
explored its interaction with three terms using appropriate centering (Brincks et al.
2017; Dalal and Zickar 2012; Enders and Tofighi 2007): namely with student
gender, immigrant background and socioeconomic background. None of these
terms showed a significant effect and were removed from the reported model. We
also included a product term between the open classroom discussion level of
schools and the socioeconomic intake of schools, with both covariates centered at
the country levels. The model can be expressed using Eqs. (6.1)–(6.3), which are
specified for each of the three response variables being studied, namely support for
equal rights for immigrants, different ethnic groups and women, as described in
Chap. 2:



Yijk ¼ p0jk þ p1jk xijk  x:jk þ p2jk mijk  m  :jk


ð6:1Þ
þ p3jk wijk  w  :jk þ p4jk Zijk  Z::k þ eijk



p0jk ¼ b00k þ b01k x:jk  x::k þ b02k mijk  m  :jk




þ b03k w :jk  w ::k þ b04k w  ::k  x:jk  x::k
 :jk  w ð6:2Þ
þ r0jk

b00k ¼ c000 þ m00k ð6:3Þ

In Chap. 5, we used a general equation form to express the estimated models

(Eqs. 5.1–5.3). However, in this chapter, we provide further details, in order to
explicitly state the role of centering of our variables on the interaction between the
socioeconomic status (SES) of the school intake and schools differences in the open
classroom discussion scores within each country. Here Y stands for the outcome
variables, xijk for student socioeconomic background (NISB), mijk for student civic
knowledge scores (PV1CIV–PV5CIV) divided by ten, wijk for student rates of open
classroom discussion, and Zijk for the two control variables, namely gender
(SGENDER, 0 = boy, 1 = girl) and student immigrant background (IMMIG,
0 = non-immigrant, 1 = immigrant background).
94 D. Carrasco and D. Torres Irribarra

To estimate model results, we fitted a series of multilevel models using Mplus v7

(Muthén and Muthén 2012); multilevel pseudo maximum likelihood accounted for
sampling design and scaling weights to sample size (Asparouhov 2006; Snijders
and Bosker 2012). Changing the scaling methods of the weights had little effect on
the results.1 Civic knowledge plausible values were all included in the model, and
estimates were appropriately combined (Rutkowski et al. 2010).

6.4 Results

6.4.1 Overall Fit

Each of the estimated models present a better fit in comparison to their nested

counterpart (see Table 6.2). We compared each estimated model by means of their
deviances (−2LL), Akaike information criterion (AIC), and Bayesian information
criterion (BIC). Since Mplus estimates one model for each of the plausible values,
each fit index presents a mean point estimate and a standard deviation for each
estimation. As comparing all the models by −2LL, AIC and BIC reached the same
general conclusions, here we describe the relative comparison of AIC and BIC
indexes alone. The general sequence of models starts from the null model, where all
selected covariates were fixed to zero, and progresses to the most complex model,
the moderation model, where selected covariates were allowed to vary. If AIC and
BIC reach lower values, in contrast to a nested model, the most complex model is
preferred. The null model was compared to the control model, where only the
control variables (socioeconomic status, civic knowledge, gender, and immigrant
background) were allowed to vary. This comparison favored the control model. The
next or main model, which additionally included open classroom discussion,

Table 6.2 Fit statistics

Criterion for model Model
selection Null Control Main Moderation
−2 LL 2353257.59 2330616.86 2327466.02 2327410.05
(0.00) (32.75) (29.33) (29.41)
AIC 2353293.59 2330688.86 2327550.02 2327500.05
(0.00) (65.50) (58.65) (58.81)
BIC 2353400.23 2330902.14 2327798.84 2327766.65
(0.00) (65.50) (58.65) (58.81)
df 18 36 42 45
−2LL deviance, AIC akaike information criterion, BIC Bayesian information criterion, df degrees
of freedom. The mean standard deviation for each estimation is provided in brackets

1
Reported results were robust to changes in the scaling methods of the weights. Differences were
observed only to the third decimal point, and these were only of one unit.
6 The Role of Classroom Discussion 95

Table 6.3 Random effects estimates, multilevel model

Dependent variable Parameter E SE P
Gender equality Intercept 50.52 −0.63 0.00
Within variance 69.08 −3.73 0.00
Between school variance 2.13 −0.28 0.00
Between country variance 13.84 −3.52 0.00
Ethnic equality Intercept 50.80 −0.60 0.00
Within variance 73.69 −3.64 0.00
Between school variance 2.44 −0.26 0.00
Between country variance 10.42 −3.09 0.00
Immigrant equality Intercept 51.39 −0.56 0.00
Within variance 75.83 −3.91 0.00
Between school variance 3.16 −0.40 0.00
Between country variance 11.76 −2.31 0.00
E estimated coefficients; SE standard deviation; P p-value

compared favorably with the control model. Finally, the most complex moderation
model, which included interaction terms, open classroom discussion school means
and socioeconomic status school means, also fitted the data better than its nested
counterpart (Table 6.2). Overall, the relative fit of the models favored our selection
of variables. The intra-class correlation coefficient at the school level was in the
range 5.2–5.1% for each outcome, whereas the intra-class correlation at the country
level, was 10–14%; most of the variance in the outcomes was thus at the student
level (Table 6.3).

6.4.2 Main Effects

Overall, schools with higher levels of open classroom discussion had students who
were more likely to endorse gender equality (b03k = 0.20, SE = 0.02, p < 0.01),
hold higher levels of support for equal rights for all ethnic groups (b03k = 0.21,
SE = 0.02, p < 0.01), and show greater support for equal rights for immigrants
(b03k = 0.18, SE = 0.02, p < 0.01). While the control variables accounted for 52,
44 and 34% of the variance between schools for each respective outcome, adding
schools’ open classroom discussion levels accounts for 7, 8 and 5% additional
variance for each outcome, respectively.
School composition, in terms of socioeconomic background and levels of civic
knowledge, also showed positive relationships between schools. School environments
with a higher proportion of students with a higher socioeconomic background
displayed higher mean levels of support for gender equality (b01k = 0.53,
SE = 0.13, p < 0.01), higher levels of support for equal rights for all ethnic groups
(b01k = 0.29, SE = 0.15, p < 0.01), and greater support for equal rights for immigrants
(b01k = 0.36, SE = 0.14, p < 0.01). Similarly, schools with higher levels of civic
knowledge also showed higher levels of endorsement for equal rights for women
96 D. Carrasco and D. Torres Irribarra

(b02k = 0.23, SE = 0.02, p < 0.01), all ethnic groups (b02k = 0.21, SE = 0.12,
p < 0.01), and immigrants (b02k = 0.14, SE = 0.02, p < 0.01). However, these
differences were not attributable to school contextual effects; that is, they were not
attributable to the unique school contribution to these relationships.

6.4.3 Moderation Effects

There was a negative interaction between school open classroom discussion levels
and school socioeconomic levels, relative to support for equal rights for women
(b04k = −0.07, SE = 0.02, p < 0.01). A negative coefficient implies a buffer effect:
a school’s intake is positively related to the higher endorsement of gender equality,
yet conditional on the level of open classroom discussion within schools (see
Table 6.4). Thus, schools with a high intake of students from lower socioeconomic
backgrounds, yet with higher than average open classroom discussion, are expected
to have a higher level of endorsement for gender equality than other similar schools
with lower levels of open classroom discussion. To assess these findings, we fitted
the same implied model for each country. This enabled us to assess the consistency
Table 6.4 Fixed effects estimates, multilevel model
Variables Within school Between school
estimates estimates
Dependent Independent E SE P E SE P
Gender equality SES 0.18 0.05 0.00 0.53 0.13 0.00
Civic knowledge 0.34 0.01 0.00 0.23 0.02 0.00
Gender 2.98 0.35 0.00
Immigrant background 1.08 0.24 0.00
Open classroom discussion 0.13 0.01 0.00 0.20 0.02 0.00
SES: open classroom −0.07 0.02 0.00
discussion
Ethnic equality SES 0.20 0.06 0.00 0.29 0.15 0.06
Civic knowledge 0.31 0.01 0.00 0.21 0.02 0.00
Gender 1.81 0.21 0.00
Immigrant background 3.49 0.66 0.00
Open classroom discussion 0.14 0.01 0.00 0.20 0.02 0.00
SES: open classroom 0.01 0.02 0.76
discussion
Immigrant SES 0.18 0.07 0.01 0.36 0.14 0.01
equality Civic knowledge 0.24 0.01 0.00 0.14 0.02 0.00
Gender 1.76 0.24 0.00
Immigrant background 5.08 0.64 0.00
Open classroom discussion 0.14 0.01 0.00 0.20 0.02 0.00
SES: open classroom −0.01 0.03 0.67
discussion
SES socioeconomic status; E estimated coefficients; SE standard deviation; P p-value
6 The Role of Classroom Discussion 97

Fig. 6.1 Interaction effect between open classroom discussion scores and the average socio-
economic level of school intake on support for equal rights for women. Unstandardized
coefficients for the interaction term of open classroom discussion school means and SES school
means. Mean estimates are plotted as black dots, with accompanying lines indicating the extent of
the 95% confidence intervals. Results from Liechtenstein are not included, as these were beyond
acceptable confidence limits. The mean for all countries is indicated by a dotted line

of our results, given that pooled coefficients may be “overpowered” by the size of
the samples involved in these estimates. Results by country showed that the
moderation effect was not a consistent estimate for all countries (see Fig. 6.1). The
98 D. Carrasco and D. Torres Irribarra

results of the single-country models indicated that a statistically significant inter-

action between school open classroom discussion and school socioeconomic status
was only found for Austria. Austria was thus the only country where schools with
similar socioeconomic intakes reported stronger support for women’s equal rights
when there was a greater level of open classroom discussion.

6.5 Discussion and Conclusions

School practices for the discussion of controversial issues are important for students
and school egalitarian attitudes. The levels of openness to the discussion of political
and social issues in classrooms during regular lessons were systematically related to
student attitudes toward equal rights for women, all ethnic groups and immigrants.
This relationship is positive when pooled across all jurisdictions. By partitioning
student scores of perceptions of openness in classroom discussion into school
means and student deviations from school means, we were able to examine the role
of this learning environment factor (Lüdtke et al. 2009). These patterns of results
were robust when controlling for student characteristics, such as gender, immigrant
background, socioeconomic background and student civic knowledge. They were
also unaffected by school differences in terms of school socioeconomic intake and
the overall civic knowledge of students in school.
What “schools do” matters in establishing students’ support for equal rights. The
general idea, that social attitudes, such as prejudice, racism and sexism are learned
and developed also leads to the idea that these attitudes may be unlearned (Zick
et al. 2011). Relevant school climate factors suggest potential school differences
that may foster the development of egalitarian attitudes toward others. Openness to
discussion in a school may not only be important for its relation to civic knowledge
(Schulz 2002; Schulz et al. 2010; Torney et al. 1975), it may also establish interest
in informed voting and the ability to embrace conflict within democracy (Campbell
2008; Godfrey and Grayman 2014). In the light of the results in this chapter, open
classroom discussion may also be important for fostering egalitarian attitudes among
students. Van Hiel et al. (2004) suggested that educational interventions aimed at
reducing the “need for closure”, a form of cognitive closed-mindedness, might reduce
authoritarianism, a common predictor of prejudice. School interventions with teachers
have been able to promote higher levels of open classroom discussion in the United
States (Barr et al. 2015). However, these have not translated into a reduction of
prejudice. Current results are encouraging, however, showing positive results for this
line of reasoning across different contexts.
Discussion of political and social issues within classrooms is often avoided in
schools (Quaynor 2012). Encouraging students to discuss controversial issues and
allowing them to make up their own minds, while presenting several sides of the
argument, requires a teacher who displays committed impartiality (Kelly 1986);
teachers are not only required to balance classroom discussion to be inclusive of
different views but also participate in the discussions with a personal position on the
6 The Role of Classroom Discussion 99

issue. Without proper institutional support for teachers by local school authorities,
discussing controversial issues involving race, immigration and gender in the
classroom may be silenced by self-censorship. Regional and national perspectives
regarding the gender rights, institutional discrimination between races, and immi-
gration may establish that large differences exist regarding what are the current
norms and how far these are from ethical ideals of equal rights for all. Thus, clear
curricular guidelines and support for teachers can be powerful tools to encourage
classroom discussion of political and social issues as a common school practice,
and through it fostering improved political attitudes and civic engagement.

References

Asparouhov, T. (2006). General multi-level modeling with sampling weights. Communications in

Statistics: Theory and Methods, 35(3), 439–460. https://doi.org/10.1080/03610920500476598.
Barber, C., Sweetwood, S. O., & King, M. (2015). Creating classroom-level measures of
citizenship education climate. Learning Environments Research, 18(2), 197–216. https://doi.
org/10.1007/s10984-015-9180-7.
Barr, D. J., Boulay, B., Selman, R. L., Mccormick, R., Lowenstein, E., Gamse, B., et al. (2015).
A randomized controlled trial of professional development for interdisciplinary civic education:
Impacts on humanities teachers and their students. Teachers College Record, 117(20307), 1–
52. Retrieved from http://www.tcrecord.org/content.asp?contentid=17470.
Brincks, A. M. (2012). The implications of centering in a three-level multilevel model. PhD thesis,
University of Miami, FL. Open Access Dissertations, 743. Miami, FL: University of Miami.
Retrieved from http://scholarlyrepository.miami.edu/cgi/viewcontent.cgi?article=1753&context=
oa_dissertations.
Brincks, A. M., Enders, C. K., Llabre, M. M., Bulotsky-Shearer, R. J., Prado, G., & Feaster, D.
J. (2017). Centering predictor variables in three-level contextual models. Multivariate
Behavioral Research, 52(2), 149–163. https://doi.org/10.1080/00273171.2016.1256753.
Campbell, D. E. (2008). Voice in the classroom: How an open classroom climate fosters political
engagement among adolescents. Political Behavior, 30(4), 437–454. https://doi.org/10.1007/
s11109-008-9063-z.
Caro, D., & Lenkeit, J. (2012). An analytical approach to study educational inequalities: 10
hypothesis tests in PIRLS 2006. International Journal of Research and Method in Education,
35(1), 3–30. https://doi.org/10.1080/1743727X.2012.666718.
Caro, D., & Schulz, W. (2012). Ten hypotheses about tolerance toward minorities among Latin
American adolescents. Citizenship, Social and Economics Education, 11(3), 213–234. https://
doi.org/10.2304/csee.2012.11.3.213.
Carvacho, H., Zick, A., Haye, A., González, R., Manzi, J., Kocik, C., et al. (2013). On the relation
between social class and prejudice: The roles of education, income, and ideological attitudes.
European Journal of Social Psychology, 43(4), 272–285. https://doi.org/10.1002/ejsp.1961.
Coenders, M., & Scheepers, P. (2003). The effect of education on nationalism and ethnic
exclusionism: An international comparison. Political Psychology, 24(2), 313–343. https://doi.
org/10.1111/0162-895X.00330.
Dalal, D. K., & Zickar, M. J. (2012). Some common myths about centering predictor variables in
moderated multiple regression and polynomial regression. Organizational Research Methods,
15(3), 339–362. https://doi.org/10.1177/1094428111430540.
Dewey, J. (1916). Democracy and education. New York, NY: The Macmillan Company.
Easterbrook, M. J., Kuppens, T., & Manstead, A. S. R. (2015). The education effect: Higher
educational qualifications are robustly associated with beneficial personal and socio-political
outcomes. Social Indicators Research. https://doi.org/10.1007/s11205-015-0946-1.
100 D. Carrasco and D. Torres Irribarra

Enders, C. K., & Tofighi, D. (2007). Centering predictor variables in cross-sectional multilevel
models: A new look at an old issue. Psychological Methods, 12(2), 121–138. https://doi.org/
10.1037/1082-989X.12.2.121.
Godfrey, E. B., & Grayman, J. K. (2014). Teaching citizens: The role of open classroom climate in
fostering critical consciousness among youth. Journal of Youth and Adolescence, 43(11),
1801–1817. https://doi.org/10.1007/s10964-013-0084-5.
Heck, R. H., & Thomas, S. L. (2015). An introduction to multilevel modeling techniques: MLM
and SEM approaches Using MPLUS. New York, NY and Hove, UK: Routledge.
Highton, B. (2009). Revisiting the relationship between educational attainment and political
sophistication. The Journal of Politics, 71(4), 1564. https://doi.org/10.1017/
S0022381609990077.
Hooghe, M., Meeusen, C., & Quintelier, E. (2013a). The impact of education and intergroup
friendship on the development of ethnocentrism. A latent growth curve model analysis of a
five-year panel study among Belgian late adolescents. European Sociological Review, 29(6),
1109–1121. https://doi.org/10.1093/esr/jcs086.
Hooghe, M., Verhaegen, S., & Quintelier, E. (2013b). The relationship between political trust,
generalized trust and European identity among adolescents. In ECPR General Conference
(pp. 1–26). Bordeaux. Retrieved from https://ecpr.eu/filestore/paperproposal/e931616a-db49-
40f1-8e74-e22cdb42531e.pdf.
Houtman, D. (2003). Lipset and “working-class” authoritarianism. The American Sociologist, 34
(1–2), 85–103. https://doi.org/10.1007/s12108-003-1008-8.
Kelly, T. E. (1986). Discussing controversial issues: Four perspectives on the teacher’s role.
Theory and Research in Social Education, 14(2), 113–138. https://doi.org/10.1080/00933104.
1986.10505516.
Lenzi, M., Vieno, A., Sharkey, J., Mayworm, A., Scacchi, L., Pastore, M., et al. (2014). How
school can teach civic engagement besides civic education: The role of democratic school
climate. American Journal of Community Psychology, 54(3–4), 251–261. https://doi.org/10.
1007/s10464-014-9669-8.
Lipset, S. M. (1959). Democracy and working-class authoritarianism. American Sociological
Review, 24(4), 482–501. Retrieved from http://www.jstor.org/stable/2089536.
Lüdtke, O., Marsh, H. W., Robitzsch, A., Trautwein, U., Asparouhov, T., & Muthén, B. (2008).
The multilevel latent covariate model: a new, more reliable approach to group-level effects in
contextual studies. Psychological Methods, 13(3), 203–229. https://doi.org/10.1037/a0012869.
Lüdtke, O., Robitzsch, A., Trautwein, U., & Kunter, M. (2009). Assessing the impact of learning
environments: How to use student ratings of classroom or school characteristics in multilevel
modeling. Contemporary Educational Psychology, 34(2), 120–131. https://doi.org/10.1016/j.
cedpsych.2008.12.001.
Luskin, R. C. (1990). Explaining political sophistication. Political Behavior, 12(4), 331–361.
https://doi.org/10.1007/BF00992793.
McNeish, D., Stapleton, L. M., & Silverman, R. D. (2017). On the unnecessary ubiquity of
hierarchical linear modeling. Psychological Methods, 22(1), 114–140. https://doi.org/10.1037/
met0000078.
Muthén, L. K., & Muthén, B. (2012). Mplus user’s guide (7th ed.). Los Angeles, CA: Muthén &
Muthén. Retrieved from http://www.statmodel.com/.
O’Connell, A. A., & McCoach, D. B. (2008). Multilevel modeling of educational data. Charlotte,
NC: IAP.
Quaynor, L. J. (2012). Citizenship education in post-conflict contexts: A review of the literature.
Education, Citizenship and Social Justice, 7(1), 33–57. https://doi.org/10.1177/
1746197911432593.
Rutkowski, L., Gonzalez, E., Joncas, M., & von Davier, M. (2010). International large-scale
assessment data: Issues in secondary analysis and reporting. Educational Researcher, 39(2),
142–151. https://doi.org/10.3102/0013189X10363170.
Schulz, W. (2002). Explaining differences in civic knowledge: Multilevel regression analysis of
student data from 27 countries. American Educational Research Association, New Orleans,
6 The Role of Classroom Discussion 101

(April), 1–5. Retrieved from http://iccs.acer.edu.au/uploads/File/papers/AERA2001_SchulzW_

CIVEDCivicKnowledge.pdf.
Schulz, W., Ainley, J., Fraillon, J., Kerr, D., & Losito, B. (2010). ICCS 2009 international report:
Civic knowledge, attitudes, and engagement among lower-secondary school students in 38
countries. Amsterdam, The Netherlands: International Association for the Evaluation of
Educational Achievement (IEA).
Snijders, T. A. B., & Bosker, R. J. (2012). Multilevel analysis: An introduction to basic and
advanced multilevel modeling (2nd ed.). London, UK: SAGE Publications Ltd.
Stapleton, L. M., Yang, J. S., & Hancock, G. R. (2016). Construct meaning in multilevel settings.
Journal of Educational and Behavioral Statistics, 41(5), 481–520. https://doi.org/10.3102/
1076998616646200.
Torney, J. V., Oppenheim, A. N., & Farnen, R. F. (1975). Civic education in ten countries: An
empirical study. New York, NY: Wiley.
Torney-Purta, J. V. (2009). International psychological research that matters for policy and
practice. American Psychologist, 64(8), 825–837. https://doi.org/10.1037/0003-066X.64.8.825.
Van der Ploeg, P. (2016). Dewey versus “Dewey” on democracy and education. Education,
Citizenship and Social Justice, 11(2), 145–159. https://doi.org/10.1177/1746197916648283.
Van Hiel, A., Pandelaere, M., & Duriez, B. (2004). The impact of need for closure on conservative
beliefs and racism: Differential mediation by authoritarian submission and authoritarian
dominance. Personality and Social Psychology Bulletin, 30(7), 824–837. https://doi.org/10.
1177/0146167204264333.
Vollebergh, W. (1996). The development of authoritarianism in adolescence: Longitudinal change
and the impact of age, gender and educational level. In K. Hurrelmann & S. Hamilton (Eds.),
Social problems and social contexts in adolescence: Perspectives across boundaries (pp. 325–
351). New York, NY: Aldine de Gruyter.
Willms, J. D. (2010). School composition and contextual effects on student outcomes. Teachers
College Record, 112(4), 1008–1037.
Zick, A., Küpper, B., Hövermann, A., German, C., & Fenn, E. (2011). Intolerance, prejudice and
discrimination. A European report. Berlin, Germany: Nora Langenbacher
Friedrich-Ebert-Stiftung Forum Berlin Projekt. Retrieved from http://library.fes.de/pdf-files/
do/07908-20110311.pdf.

Open Access This chapter is licensed under the terms of the Creative Commons
Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/
4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in
any medium or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.